44(3):293-298,2003

FORENSIC SCIENCES

Identification of Human Remains by Immobilized Sequence-specific Oligonucleotide Probe Analysis of mtDNA Hypervariable Regions I and II Matthew N. Gabriel, Cassandra D. Calloway1, Rebecca L. Reynolds1, Dragan Primorac2 San Francisco Police Department, Criminalistics Laboratory, San Francisco; 1Roche Molecular Systems, Department of Human Genetics, Alameda, CA, USA; and 2Split University Hospital and School of Medicine, Laboratory for Clinical and Forensic Genetics, Split, Croatia

Aim. A rapid analysis of mitochondrial DNA (mtDNA) sequences with an array of immobilized sequence-specific oligonucleotide (SSO) probes was tested on 18 skeletal elements recovered from mass graves in Croatia, which could not be genotyped with common forensic nuclear DNA systems (PM+DQA1 and short tandem repeat analysis). Methods. We used duplex polymerase chain reaction (PCR) amplification of the mtDNA hypervariable regions I and II (HVI and HVII) (444 bp and 415 bp amplicons, respectively) and subsequent linear array typing, which targets six polymorphic regions and two additional sites within the human mtDNA HVI and HVII. The remaining amplified products were subjected to direct sequence analysis to obtain complete sequence information for the targeted HV regions. Result. Duplex PCR amplification of the mtDNA HVI and HVII was successful in providing sufficient product for typing with the array of SSO probes in 14 out of the 18 sample extracts. We report here the sequence match of one set of remains with a panel of immobilized SSO probes, followed by direct sequence analysis. The corresponding mtDNA haplotype obtained for the bone sample and the putative maternal reference was unique in a database of 105 randomly selected Croatian individuals. Conclusion. Mitochondrial DNA typing with an array of immobilized SSO probes can be a benefit to forensic DNA analysis of mass disaster remains and identity testing of single and mass graves. Key words: Croatia; DNA, mitochondrial; forensic anthropology; forensic medicine; oligonucleotide probes

Analysis of human mitochondrial DNA (mtDNA) control region sequences is a useful tool for forensic identity testing on a range of samples (1-3) due to the high sequence variability (4), genome copy number per cell (5), and maternal inheritance (6). Sequence variation within the human mtDNA control region has been widely established as a valuable marker for individual identification (7-11). Polymerase chain reaction (PCR)-based techniques used to detect variation include direct DNA sequence analysis and screening with a panel of sequence-specific oligonucleotide (SSO) probes (3,9,10,12-20). These studies have established a high degree of polymorphism located primarily within two regions of the human mtDNA control region known as hypervariable regions I and II (HVI and HVII). The high mtDNA genome copy number per cell (~500-1,000 copies) allows analysis of the most challenging forensic specimens, such as shed hairs from crime scenes and significantly aged remains (1-4,17,21-24). For this reason, mtDNA testing is often successful in cases where nuclear markers

cannot be amplified. In addition, the maternal inheritance of mtDNA genomes allows a greatly expanded reference sample population for human identification efforts. Typing of human mtDNA sequences by using SSO probes has been described elsewhere (12,17, 25,30). Reynolds et al (17) reported the detection of sequence variation for 689 unrelated individuals, using a typing strip that contained an array of immobilized SSO probes specific for variants within five polymorphic regions of HVII. This method utilizes immobilized SSO probes for rapid sample screening in a reverse line blot format. From this study, the panel of 16 immobilized SSO probes for HVII was shown to provide a significant power of discrimination, ranging from ~0.92-0.98, among African American, US Caucasian, US Hispanic, and Japanese populations (17). The original typing strip has since been expanded to include four polymorphic regions and one site within HVI as well as an additional HVII site for sequence variation detection (HVI A, C, D, E, and HVII A, B, C,

www.cmj.hr 293

Gabriel et al: SSO Probe Analysis of mtDNA HVI and HVII

Croat Med J 2003;44:293-298

D, and positions 189 and 16093 relative to the Anderson reference sequence) (12,26). Two versions of the expaded HVI/HVII typing strip were used in this analysis. The first version contains 27 possible immobilized SSO probe signals (HVI A, C and HVII A, B, C, D, and 189 and 16093), whereas the fully expanded array contains 31 possible probe signals (HVI A, C, D, E and HVII A, B, C, D, and 189 and 16093) for HVI and HVII. Additional probes are currently being investigated as potential candidates for sequence variation detection. Although there is a greater degree of polymorphism within HVI than HVII, the distribution of variable positions across HVI is less clustered than HVII. Therefore, fewer candidate HVI polymorphic regions are idealfor analysis by probe hybridization methods. Despite the relatively high success rate obtained with nuclear PM+DQA1 and short tandem repeat analysis (~85%) on bone samples recovered from similar mass graves, the eighteen extracts tested here could not be amplified with nuclear markers (27). In this paper, we report a duplex PCR amplification strategy for full-length HVI and HVII amplicons of 444 bp and 415 bp, respectively, and subsequent typing of skeletal elements recovered from mass graves in Croatia (27,28).

mmol/L each primer, and 0.25 units/mL AmpliTaq Gold® DNA Polymerase (PE Biosystems). A group of 18 bone sample extracts plus negative and positive controls were prepared and amplified. A separate set of 4 potential maternal reference samples was prepared in a different PCR setup hood and after the bone sample setup to avoid the potential for cross-contamination. All reactions were performed in a PE Biosystems 9600 model thermal cycler by use of a 12-minute activation step at 92 °C. Samples were subjected to 38 cycles at 92 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s, followed by a final 72 °C extension step for 10 min. Sequences for the biotinylated HVI and HVII primers are as follows: HVI: F15975 5'-XCTCCACCATTAGCACCCAA-3' and R16418 5'-XATTTCACGGAGGATGGTG-3'; HVII: F15 5'-XC ACCCTATTAACCACTCACG-3' and R429 5'-XCTGTTAAAAG TGCATACCGCCA-3' (where X denotes biotin). PCR products were analyzed on 1.5% agarose gels in 1X Tris-Borate-EDTA electrophoresis buffer (TBE) stained with 0.5 mg/mL ethidium bromide. Products were used directly for SSO typing and then purified with Millipore Ultrafree-MC® centrifugal filter devices (Millipore, Bedford, MA, USA) to eliminate excess primer prior to sequence analysis. For filtration, 25-40 mL of PCR product was added to 350 mL of Tris-EDTA (TE), pH 8.0, and centrifuged for 4 min at 12,000 rpm. After 20 mL of TE was added to the membrane, the recovery was transferred to a clean microcentrifuge tube for cycle sequencing. Immobilized SSO Probe Design A complete description of immobilized SSO probe design has been reported elsewhere (17), and those guidelines were followed for designing additional probes in the A, C, D, and E regions of HVI. In general, the ascending order of destabilizing mismatches is as follows: A-T,G-C